Re: [Computer-go] Go Tournament with hinteresting rules

[Darren Cook]

> I have been told that bots that are based on MC play better when they only 
> record the result of each roll out (W or L)
> rather than the margin of victory.
> 
> To me this is counter-intuitive.
> 
> Does anyone have an intelligible reason why it should be so?

The search then optimizes for the probability of winning, rather than
optimizing for the largest margin of victory.

Imagine two stock traders. Their goal is to beat the market over the
next 12 months, and they both choose 10 companies from the FTSE100.
Trader A randomly chooses 10 companies with dividends that are paying
over the average for the FTSE100. Trader B chooses the 10 companies with
the highest dividends. Intuitively trader B should have earned more by
the end of the year, but there is a good chance that at least one
company will go bankrupt, and another will cut its dividend. Maybe the
other 8 choices will do well enough to keep him ahead overall, but
chances are that trade A will come out ahead.

Games of go tend to be dominated by life and death battles. There may be
a way for black to kill white's group, and win big, but it is awfully
complicated and we don't have time for exhaustive search. If we can
still win by letting white's group live small, that is a much safer path.

There is also a pragmatic reason: it is just one bit of information to
pass up the tree, so very easy to make a single number for chance of
win. With margin of victory you end up with the problem of how to pass a
probability distribution up the tree, and then what to do with it at the
top.
(The presence of the life/death battles means the distribution tends to
have multiple peaks, not be nice and gaussian.)

Darren



-- 
Darren Cook, Software Researcher/Developer
My New Book: Practical Machine Learning with H2O:
  http://shop.oreilly.com/product/0636920053170.do
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Go Tournament with hinteresting rules

[Lucas, Simon M]

The intelligible reason is that focussing on the win or loss
means that the bot is focussing on what actually matters: winning
and not losing.  If the bot focuses on the margin of victory
the play can be skewed to aim for big wins that may not
happen while paying insufficient attention to small losses.

Simon Lucas



-Original Message-
From: Computer-go [mailto:computer-go-boun...@computer-go.org] On Behalf Of 
Charles Leedham-green
Sent: 08 December 2016 23:23
To: computer-go@computer-go.org
Subject: Re: [Computer-go] Go Tournament with hinteresting rules

I have been told that bots that are based on MC play better when they only 
record the result of each roll out (W or L) rather than the margin of victory.

To me this is counter-intuitive.

Does anyone have an intelligible reason why it should be so?

Charles

> On 8 Dec 2016, at 22:56, Erik van der Werf <erikvanderw...@gmail.com> wrote:
> 
> On Thu, Dec 8, 2016 at 10:58 PM, "Ingo Althöfer" <3-hirn-ver...@gmx.de> wrote:
>> Playing under such conditions might be a challenge for the bots
> 
> Why? Do you think the humans will collude?  ;-)
> 
> Erik.
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Go Tournament with hinteresting rules

[David Fotland]

Because you test it both ways, and one wins more games. Many things about the 
playout policy are mysterious and can only be tested to see if they make play 
stronger. Often the results of testing are counterintuitive. I'd guess only 
about a quarter of the things I tried in Many Faces made the program stronger.

I can think of several possible explanations, but that's not science, it's 
telling a story with no evidence.

David

> -Original Message-
> From: Computer-go [mailto:computer-go-boun...@computer-go.org] On Behalf Of
> Charles Leedham-green
> Sent: Thursday, December 08, 2016 3:23 PM
> To: computer-go@computer-go.org
> Subject: Re: [Computer-go] Go Tournament with hinteresting rules
> 
> I have been told that bots that are based on MC play better when they only
> record the result of each roll out (W or L) rather than the margin of
> victory.
> 
> To me this is counter-intuitive.
> 
> Does anyone have an intelligible reason why it should be so?
> 
> Charles
> 
> > On 8 Dec 2016, at 22:56, Erik van der Werf <erikvanderw...@gmail.com>
> wrote:
> >
> > On Thu, Dec 8, 2016 at 10:58 PM, "Ingo Alth fer" <3-hirn-ver...@gmx.de>
> wrote:
> >> Playing under such conditions might be a challenge for the bots
> >
> > Why? Do you think the humans will collude?  ;-)
> >
> > Erik.
> > ___
> > Computer-go mailing list
> > Computer-go@computer-go.org
> > http://computer-go.org/mailman/listinfo/computer-go
> 
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Go Tournament with hinteresting rules

[Charles Leedham-green]

I have been told that bots that are based on MC play better when they only 
record the result of each roll out (W or L)
rather than the margin of victory.

To me this is counter-intuitive.

Does anyone have an intelligible reason why it should be so?

Charles

> On 8 Dec 2016, at 22:56, Erik van der Werf  wrote:
> 
> On Thu, Dec 8, 2016 at 10:58 PM, "Ingo Althöfer" <3-hirn-ver...@gmx.de> wrote:
>> Playing under such conditions might be a challenge for the bots
> 
> Why? Do you think the humans will collude?  ;-)
> 
> Erik.
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Go Tournament with hinteresting rules

[Michael Markefka]

>
>
>   The basic explanation for why this is not straightforward is that you
> never want your program to consider moves in the direction of
> low-probability wins, no matter how large margins they might have; the
> MC measurement function is very noisy with regards to individual samples.
>

I do wonder though whether a final score value network would work better
than MC here, and whether there could be a minimum win percentage threshold
that could work. I'd love to see someone implement a final score value
network and chose moves according to expected score or expected value
(expected winning percentage * expected final score), with a minimum filter
for expected winning percentage.
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Go Tournament with hinteresting rules

[Petr Baudis]

  Hi!

On Thu, Dec 08, 2016 at 11:23:50PM -0800, Freeman Ng wrote:
> No, it's because the bots' mc based algorithms currently don't care how
> much they win by. (At least I'm assuming that's what Ingo meant.) They just
> try to maximize their odds of winning.
> 
> I've often wondered about this, though, and maybe the bot developers here
> can give me an answer. There's no reason why an mc-based go program
> couldn't also factor winning margin into its decisions, is there? I assume
> that at some point, what the mc analysis yields is a winning probability
> for each candidate move, but at that point, you could still combine that
> number with other factors, right? Some combination of winning probability
> and probable winning margin, so that, for example, a 87% chance of winning
> by 5 points could be rated lower than a 85% change of winning by 20. I
> don't know what the ideal formula would be, and you'd probably want to
> prevent the winning probability from ever getting too low, while also
> ignoring potentially large winning margins beyond a certain point, but the
> idea would be to generally make the bots play more like humans.

  I think many programs use this hack.  (In Pachi, we eventually
disabled it because it compromised strength noticeably.)

  The basic explanation for why this is not straightforward is that you
never want your program to consider moves in the direction of
low-probability wins, no matter how large margins they might have; the
MC measurement function is very noisy with regards to individual samples.

> Do any commercial Go programs work this way? If not, I'd like to request it
> from the commercial developers here. It could be an option that you'd only
> have in the commercial product, for your users to turn on if they prefer
> it. You could still operate in pure mc mode for bot vs. bot play.

  A more robust version of this strategy, that is used commonly in many
programs, is dynamic komi.  But it can also exhibit unstable behavior
in some cases, especially in late endgame.  I didn't quite understand
why is this important for commercial programs in particular?

  (For this mode in particular, e.g. Pachi has a special mode where it
enables a variety of such features (but is weaker in general) when given
the "maximize_score" parameter.  I wonder if other programs also have
multiple modes like this.)

Petr Baudis
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Go Tournament with hinteresting rules

[Freeman Ng]

No, it's because the bots' mc based algorithms currently don't care how
much they win by. (At least I'm assuming that's what Ingo meant.) They just
try to maximize their odds of winning.

I've often wondered about this, though, and maybe the bot developers here
can give me an answer. There's no reason why an mc-based go program
couldn't also factor winning margin into its decisions, is there? I assume
that at some point, what the mc analysis yields is a winning probability
for each candidate move, but at that point, you could still combine that
number with other factors, right? Some combination of winning probability
and probable winning margin, so that, for example, a 87% chance of winning
by 5 points could be rated lower than a 85% change of winning by 20. I
don't know what the ideal formula would be, and you'd probably want to
prevent the winning probability from ever getting too low, while also
ignoring potentially large winning margins beyond a certain point, but the
idea would be to generally make the bots play more like humans.

Why, you might ask, when this would only compromise their strength? For two
reasons:

1. To compete in tournaments like the one Ingo reported in this thread.
2. To make commercial programs more pleasing and useful for humans to play.

On #2, I don't know about others, but I really hate playing against Go
software because of how it starts playing "bad" moves once it's winning,
and I'd also like to use its winning margins to help gauge my strength.

Do any commercial Go programs work this way? If not, I'd like to request it
from the commercial developers here. It could be an option that you'd only
have in the commercial product, for your users to turn on if they prefer
it. You could still operate in pure mc mode for bot vs. bot play.

Freeman


*_*
*|| Freeman Ng - www.AuthorFreeman.com *
*||**Poetry Pro-Am - connecting our students with our poets*
*||**Who Am I? - a personalizable picture book*
*||**Haiku Diem - a daily haiku feed *
*||**Joan - a novel of Joan of Arc*
***



On Thu, Dec 8, 2016 at 2:56 PM, Erik van der Werf 
wrote:

> On Thu, Dec 8, 2016 at 10:58 PM, "Ingo Althöfer" <3-hirn-ver...@gmx.de>
> wrote:
> > Playing under such conditions might be a challenge for the bots
>
> Why? Do you think the humans will collude?  ;-)
>
> Erik.
>
>
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Go Tournament with hinteresting rules

[Marc Landgraf]

Well, the system i identical to that besides:
- there are 19 bonus points for winning (I really like that one...)
- it is capped at +40 an -40

But I do not think it is too interesting for bots right now mostly due to
lack of similar strengths bots. And while the GtI tourney does equalize
this with handicap, the way bots deal with handi, especially when being of
different strengths, kinda ruins it.

2016-12-09 0:11 GMT+01:00 Lukas van de Wiel :

> So why not add the amount of points equal to your score? You win by 16.5
> points? You get 16.5 points. You lose by 4.5. You lose 4.5 points. At the
> end of the tournament, there will be contestants with a negative score, but
> it seems a more straightforward system to me, and the players losing only
> by a small amount will still rate righter than those that are slaughtered.
>
> On Thu, Dec 8, 2016 at 11:56 PM, Erik van der Werf <
> erikvanderw...@gmail.com> wrote:
>
>> On Thu, Dec 8, 2016 at 10:58 PM, "Ingo Althöfer" <3-hirn-ver...@gmx.de>
>> wrote:
>> > Playing under such conditions might be a challenge for the bots
>>
>> Why? Do you think the humans will collude?  ;-)
>>
>> Erik.
>> ___
>> Computer-go mailing list
>> Computer-go@computer-go.org
>> http://computer-go.org/mailman/listinfo/computer-go
>>
>
>
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Go Tournament with hinteresting rules

[Lukas van de Wiel]

So why not add the amount of points equal to your score? You win by 16.5
points? You get 16.5 points. You lose by 4.5. You lose 4.5 points. At the
end of the tournament, there will be contestants with a negative score, but
it seems a more straightforward system to me, and the players losing only
by a small amount will still rate righter than those that are slaughtered.

On Thu, Dec 8, 2016 at 11:56 PM, Erik van der Werf  wrote:

> On Thu, Dec 8, 2016 at 10:58 PM, "Ingo Althöfer" <3-hirn-ver...@gmx.de>
> wrote:
> > Playing under such conditions might be a challenge for the bots
>
> Why? Do you think the humans will collude?  ;-)
>
> Erik.
> ___
> Computer-go mailing list
> Computer-go@computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] Go Tournament with hinteresting rules

[Erik van der Werf]

On Thu, Dec 8, 2016 at 10:58 PM, "Ingo Althöfer" <3-hirn-ver...@gmx.de> wrote:
> Playing under such conditions might be a challenge for the bots

Why? Do you think the humans will collude?  ;-)

Erik.
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

[Computer-go] Go Tournament with hinteresting rules

[Ingo Althöfer]

Hi,
this is info on a human-human go tournament, but with
interesting scoring.
Wins by 40 or more points (or by resignation) give score 100,
Win by 39 gives score 99
Win by 38 gives score 98 ...
Win by 1 gives score 60.

Loss by 40 or more gives score 0,
Loss by 39 gives score 1
...
Loss by 1 gives score 40

The tournament was high prized (for European Go standards).
It toook place in Berlin three weks ago:
http://eurogofed.org/index.html?id=88

Playing under such conditions might be a challenge for the bots
based on MC.

Ingo.


___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go